Complexity in Government

Today was the first day of my three day programme at the Civil Service College in Singapore. It’s my second year as Senior Visiting Fellow and I am really enjoying it. To see a whole government start to shift towards complexity theory as a new paradigm of government is mist encouraging. On this assignment I am working over two days with a small group of selected civil servants to define a set of experimental projects. I’ve also got my usual meetings (and intellectual stimulation) with the Head of the Civil Service and others.

Today was a presentation to a wider audience before moving on to the detailed work. I started with the Handy S Curve models with my most elaborate presentation to date (slides are here, plus a 117mb podcast). My argument was that we in a phase shift period, between the paradigm of systems dynamics (process, scenario planning, learning organisation) and complexity theory (fragmented sense-making, use of human sensor networks, the inter-connectivity of social computing etc.). In such a period its essential to do a few things:

• Don’t abandon past practice, but realise the limitations. Process control works for ordered aspects of systems, but only for order. Scenario Planning has high utility to create awareness of possibilities, but needs better generative methods and is poor for monitoring. Complexity theory is not a universal all embracing management fad, its a way to understand the boundaries if applicability (the Cynefin framework focuses on that) and also an approach to un-ordered and ambiguous situations.
• During a period of change you can’t work from past practice, organisations that do go under or undergo some form of catastrophic failure from which they may or may not recover. I used the example of IBM here, one of the supreme dominant predators in the functional period, who failed to make the transition until a massive lost forced significant change. The approach therefore is theory informed practice, applying sound theory (which for me means natural science in the main) and monitoring for emergence.
• You can’t put new wine into old wineskins, new methods and tools have to be adopted in full, attempting to dress them up in the familiar clothes of the previous paradigm helps no one. All you will get is failure and set back. Now that doesn’t mean that complexity based projects are high risk, in part because done well they always comprise small scale safe-fail interventions mainly focused on intractable problems that conventional methods have failed to resolve.

More on this tomorrow as I start the detailed work. I have also finished writing up the Handy S-Curve material for an article (and the book) which will be available shortly. I’m going to publish a table from that tomorrow for collaborative brainstorming! I’ll also catch up on various promised blogs on how government should respond to the use of complexity as promised some months ago before I got submerged in projects.